Regional Science Policy and the Growth of Knowledge Megacentres in Bioscience Clusters Philip Cooke, Director, Centre for Advanced Studies, University of Wales, Cardiff Centre for Advanced Studies University of Wales, Cardiff 44-45 Park Place Cardiff CF10 3BB cookepn@cf.ac.uk Presented at the Regional Science Association 42nd EUROPEAN CONGRESS Dortmund, Germany, August 27-31, 2002

2

Regional Science Policy and the Growth of Knowledge Megacentres in Bioscience Clusters Abstract Changes in epistemology in biosciences are generating important spatial effects. The most notable of these is the emergence of a few ‘Bioscience Megacentres’ of basic and applied bioscience (molecular, post-genomic, proteomics, etc.) medical and clinical research, biotechnology research, training in these and related fields, academic entrepreneurship and commercial exploitation by clusters of ‘drug discovery’ start-up and spin-off companies, along with specialist venture capital and other innovation system support services. Large pharmaceutical firms that used to lead such knowledge generation and exploitation processes are becoming increasingly dependent upon innovative drug solutions produced in such clusters, and Megacentres are now the predominant source of such commercial knowledge. ‘Big pharma’ is seldom at the heart of Megacentres such as those the paper will argue are found in about four locations each in the USA and Europe, but remains important for some risk capital (‘milestone payments’), marketing and distribution of drugs discovered. The reasons for this shift (which is also spatial to some extent) are as follows: first, bioscientific research requires the formation of ‘collaboratory’ relationships among hitherto cognitively dissonant disciplines – molecular biology, combinatorial chemistry, high throughput screening, genomics, proteomics and bioinformatics to name a few. Second, the canonical ‘chance discovery’ model of bioscientific research is being replaced by ‘rational drug design’ based on those technologies because of the need massively to reduce search costs and delivery timeframes. Third, the US and to some extent European ‘Crusade against Cancer’ and other pathologies has seen major increases in basic research budgets (e.g. to $27.3 billion in 2003 for the US National Institutes of Health) and foundation expenditure (e.g. $1billion in 2003 by the UK’s Wellcome Trust; $1 billion approximately by the top ten US medical foundations, and a comparable sum from corporate foundations). Each of these tendencies weakens the knowledge generation role of ‘big pharma’and strengthens that of Megacentres. But the process also creates major, new regional disparities, which some regional governances have recognised, causing them to develop responsibilities for regional science policy and funding to offset spatial biases intrinsic in traditional national (and in the EU, supranational) research funding regimes. Responses follow a variety of models ranging from market following to both regionalised (decentralising by the centre) and ‘regionalist’ (ground-up), but in each case the role of Megacentres is justified in health terms. But their role in assisting fulfilment of regional economic growth visions is also clearly perceived and pronounced in policy terms.

3

Introduction The crucial matter of interest to this paper is the rapid decline in the capabilities of

large pharmaceuticals companies (‘big pharma’) to develop in-house new therapeutic

drug treatments, particularly those deriving from biotechnology, compared to the

rapid rise in that precise capability on the part of networks of small, dedicated

biotechnology firms (DBFs). This is commented upon in Orsenigo et al., (2001) but

these authors remain reluctant to see the facts they observe as a weakness of ‘big

pharma’. Rather, the latter is seen retaining power through its control over the former

through financing R&D contracts with milestone payments and licensing agreements,

managing due diligence, and marketing and distributing final treatments or drugs. In

this contribution we will present the shift in tacit and exploration knowledge to DBFs

as signifying a crisis for multinational drug companies. For while some other kinds of

multinational corporations adopted a strategy of downsizing central laboratories and

decentralising R&D both to branches and to the supply chain (e.g. Dupont’s major

reduction in its central Wilmington facility; GE’s restructuring of Schenectady’s role;

and the fact that aerospace firms like Bombardier of Canada routinely buy all R&D

from a globally dispersed market; Niosi, 2002), such attenuation of the R&D function

has not been sought by pharmaceuticals firms, but rather represents a failure to deal

with a thoroughgoing paradigm shift. The paper will examine the nature of that shift

and explore ways in which some ‘big pharma’ is seeking to manage its response.

The paper has four key sections:

• The first examines the knowledge management mechanisms by which DBFs

tackle the R&D or ‘drug discovery’ process and determine the nature of their

specific advantage,

• The second assesses the adequacy of these mechanisms and how industry and

intermediaries judge they need to be strengthened,

• The third examines the future of big pharma concerning the cognitive

paradigm shift linking ‘Mode 2’ knowledge production (Gibbons et al., 1994),

the demise of ‘discovery’ methods and rise of ‘rational drug design’, and fine

chemistry versus molecular biology,

4

• The fourth investigates regional development and management control issues

arising from clustering of advanced bioscientific knowledge exploration and

exploitation in a few globally significant ‘megacentres’.

Reference will also be made to a previous paper’s findings that the ‘bioscience

megacentre’ process is leading to the emergence of a new type of regional policy

called regional science policy that seeks to overcome the traditional centralising

features of nationally formulated science policies (and in the EU, supranational RTD

or research and technology policy).

2. Theoretical Approach

In the broadest terms, the theoretical approach informing the proposed paper dates at

least from Marshall (1918) and more recent transaction costs theory (Coase, 1937) as

developed by Penrose (1959) and Richardson (1972) with their ‘resource’ and

‘capabilities’ perspective on the firm, rather than the more orthodox neoclassical

theorems of Williamson (1985). This chimes also with Piore & Sabel’s (1984)

‘flexible specialisation’ conception of the advantages of small firm networks in

successfully attacking global markets. More recently authors such as Teece & Pisano

(1998) and Best (2001) have explored the ‘capabilities’ perspective in theoretical and

empirical depth.

Marshall’s initial statement was that small firms gained dynamic externalities

(nowadays more commonly referred to as ‘spillovers’, including ‘knowledge

spillovers’, see Audretsch, & Feldman, 1996; Feldman & Audretsch, 1999) from co-

location. These gave advantage in terms of specialised skills pools, opportunities for

production specialisation and technical or managerial knowledge transfer. These

circulated rapidly due to socio-cultural factors like trust, customs, social ties and other

institutional characteristics of ‘industrial districts’. Porter’s (1990; 1998) notion of

‘clusters’ owes much to these insights, as he readily admits.

However, through the twentieth century, such thinking became heterodox and other

aspects of Marshall’s contribution to economic theory, notably marginalism and

equilibrium theory, resonated better with the rise to power of the large corporation

and, within economics, theories explaining the superiority of economies of scale over

5

economies of scope (Chandler, 1990). Piore & Sabel’s (1984) discovery of a Neo-

Marshallian school of economic theory purporting to explain not the anachronistic but

globally competitive existence of modern industrial districts in Italy (e.g. Becattini,

1989; Brusco, 1989) in terms that included superior knowledge circulation and

management among small firms, caused revision to prevailing orthodoxy.

This paper will explore from a firm-capabilities and, innovatively, an institution-

capabilities perspective, the modes by which DBFs mange complex types of

knowledge ranging from basic scientific to financial in creating project networks to

generate and exploit research to develop therapeutic treatments. The extent, and

degree of formality and informality of involvement by knowledge intermediaries in

this process as compared to direct contact among peers will be explored. Moreover

the process management functions and problems of big pharma in the ‘knowledge

value chain’ from research to financing and final distribution as its control shifts

downstream will be of key theoretical interest. This is especially pertinent as a test of

the evolving thesis of transnational corporations becoming ‘hubs’ buying not making

key services (Stewart, 2001; Best, 2001). The role of knowledgeable intermediaries as

agents in innovation interactions will also be investigated to refine theory.

Theoretically there are four links connecting the paper’s key interests to the ‘design

space’ of post-genomics (Stankiewicz, 2002). First, following Polanyi (1966) and

Nonaka & Takeuchi (1995), there is the interplay of implicit (tacit) and explicit

knowledge in the project networks formed among firms with distinctive expertise, like

combinatorial chemistry, high throughput, target-based screening, genomics and

genomic libraries. The extent this mimics Nonaka & Takeuchi’s (1995) original and

Stewart’s (2001) recently reviewed ‘SECI Process’ linking ‘Socialisation’,

‘Externalisation’, ‘Creation’ and back to ‘Internalisation’ in the eliciting of implicit

knowledge, its formulation as explicit or codified knowledge, followed by its re-

internalisation as tacit knowledge can be explored. In particular, the question as to

whether there are important differences related to types of knowledge (e.g.

‘exploration’ versus ‘exploitation’) has to be confronted.

Second, how adequate are the institutional mechanisms by which such interactions are

managed? Are they largely informal and inaccurately accounted, where is formality

6

strongest, are ‘arm’s length’ or market exchanges more or less pronounced than

‘untraded interdependencies’ (Dosi, 1988)? Third, what response does big pharma

make, if any, to the shift towards ‘rational drug design’? Are there instances where in-

house capabilities to engage fully with ‘Mode 2’ knowledge production (e.g. through

acquisition, partnership alliances or attempts to mimic certain conventions more

commonly associated with DBFs, like stock options for innovative scientists or intra-

preneurship). What distinctive strategies are being pursued? Finally, what are key

barriers to control for big pharma, and what strategies are pursued to accommodate

the current deficits in in-house drug discovery? What lessons have been learned from

past experience and what new lessons are being learned currently to adjust to the

predominant knowledge value chain relationships and interactions?

Knowledge sources, including tacit/codified, internal/external, contact/intermediary,

and local/global, are key subjects of inquiry of direct relevance to regional science

and notions surrounding regional science policy. Number and type of network

partners, mechanisms for assembling partnerships, types of project and typical

expertise requirements are important to assess and compare ‘social capital’ versus

‘arm’s length’ kinds of interaction (Cooke, 2002a). To help move towards some

concrete information about this, preliminary research within such networks will be

drawn from a study being conducted in partnership with a biotechnology incubator

named Oxford BioTechNet and incubators in Germany (BioM, Munich), Israel

(Jerusalem Biotechnology Incubator), France (GenoPole, Paris) and Italy (Consorzo

Ventuno, Sardinia), based on focus-group inquiry sessions with a variety of

biotechnology project network members exploring the above and related issues in

great depth, with a view to identification of the emergent knowledge

exploration/exploitation model, its strengths and weaknesses, in comparative

perspective, and how weaknesses or gaps might be tackled. The second source of

information will be secondary information on big pharma companies. This evidence is

both quantitative and qualitative, concerning R&D performance (input-output

measures) drugs in development, drug approvals, in- licensing, out- licensing,

patenting, sub-contracting of R&D, partnerships, alliances, project-management and

knowledge-management issues. The capabilities and strategies of big pharma facing a

‘Mode 2’ knowledge production regime will thus be assessed.

7

3. The Capabilities of Dedicated Biotechnology Firms

In the ‘capabilities’ perspective on firms, and, it may be added, regions that are the

knowledge-embedded platforms in which such firms are rooted, it is dynamic

capabilities that are the most prized. This is helpful because in the literature on

‘knowledge spillovers’ it is the dynamic rather than static externalities with which

they are associated that are equally highly prized (Feldman & Audretsch, 1999).

There is an interesting debate, dating from the work of Jane Jacobs (1969) about

whether it is the diversified or specialised nature of capabilities in knowledge

spillovers that gives the basis for innovatively successful milieux. Jacobs argued in

favour of diversification, new combinations of capabilities giving rise to cognitive

progress, something with which Feldman & Audretsch broadly agree. But researchers

such as Glaeser et al. (1992) and Griliches (1992) stressed the superiority of

specialisation and the capabilities of fairly narrow ‘communities of practice’ (Seely

Brown & Duguid, 2000), known elsewhere as ‘epistemic communities’ in delving

deeply and reasonably rapidly into a particular scientific sub-field. Empirically both

showed how relatively geographically circumscribed knowledge-exploitation, for

example through patenting activity actually was. More recently though, Galison

(1997) moving well beyond the narrow confines of patenting activity such as that

relied upon by Griliches, showed convincingly that new developments in scientific

method, broadly consistent with the emergence of ‘transdisciplinarity’ in Mode 2

knowledge production (Gibbons et al., 1994) are built fundamentally on

diversification of knowledge.

This feature of contemporary ‘complexity’ in knowledge management specifically

occurring in industrial clusters is the subject of a path-breaking book edited by Curzio

& Fortis (2002). In a contribution that focuses precisely upon the issue at hand, the

following observation is made by one contributor:

‘…. innovation, and problem solving generally, depend on disciplined comparisons of alternative solutions, and these in turn require transforming tacit knowledge into what might be called pidgin formalisations: accounts sufficiently detailed to be recognisable to those who know the situations to which they refer first hand, but sufficiently abstracted from them to be accessible to outsiders, from various disciplines,’ (Sabel, 2002).

8

However, innovation is not the same as basic science. It can easily be seen that if a

scientist is collaborating with an entrepreneur, say, in writing a paper that exploits a

patent, they may or may not have to speak a kind of ‘pidgin’ scientific language to

each other. But invent ion, or discovery may well be expected to require the greater

cognitive precision associated with epistemic communities. Even Seely Brown &

Duguid (2000) who refer to the importance of communities of practice recognise,

from their long experience at Xerox PARC in Silicon Valley, that:

‘A firm, then, will almost always intersect multiple networks of practice. In Silicon Valley, for example, some firms will have cross-cutting networks of engineering, manufacturing, sales and marketing, and customer service. Networks of computer engineers, for example, will run through all the firms manufacturing computers’ (Seely Brown & Duguid, 2000)

They imply that these lateral professional links are cognitively less dissonant and

more smoothly connected even than the distinctive ‘capabilities’ linkages within the

firm. Hence the superiority of networking in a clustered environment rather than a

stand-alone competitive posture:

‘Knowledge seems to flow with particular ease where the firms involved are geographically close together. Being in the same area allowed the Apple and PARC scientists to meet and exchange ideas informally, paving the way for more formal links. Relations between PARC scientists and the Dallas engineers were in every sense far more distant’ (Seely Brown & Duguid, 2000)

Thus it seems that proximity in a cluster is fundamentally important to innovation,

that is, the stage at which deeply embedded knowledge is being confronted with

processes of knowledge exploitation and commercialisation.

When scientific method was more disciplinary than it now seems to be, that is in the

era of Mode 1 knowledge production as Gibbons et al. (1994) refer to it, it is probable

that specialisation was more important than diversification. But now there is strong

evidence, in biosciences at least, that it has become more transdisciplinary, as we have

seen, even at the exploratory R&D point in the ‘knowledge value chain’ (Cooke,

2002b) let alone the exploitation point in the same chain. Thus even basic research is

likely to contain a higher incidence of the need for ‘pidgin’ among diverse

professional scientists and engineers than used to be the case. Orsenigo et al. (2001)

date this shift in biosciences from about 1992. Thus Griliches and Glaeser were

publishing their results about specialisation at about the same time that it seems the

9

knowledge development method (epistemology) was changing significantly,

particularly in biosciences. This in turn, it can be shown related to technological

changes consequent upon precisely the rapid diversification in cognitive skills brought

about by the demands of such activities as gene sequencing and the eventual decoding

of the human genome that ushered in the post-genomic era.

Thus, in a regional ‘megacentre’ to be discussed in more detail in Section 6, that of

Northern California, it is clear that the biomedical industry relies crucially on

information and communication technologies (ICT) to decode and synthesise

bioscientific information. This is drawn from the knowledge- intensive ICT base in

Silicon Valley, and includes sequencing and screening workstations, photonics and

optical networking, low-level electrical energy instrumentation, and software among

many others. This convergence between ICT and bioscience has contributed to

discoveries in genomics, proteomics, therapeutic cloning and stem cell research, and

these in turn enable improved treatments for high ranking disease targets like cancer,

cardiovascular, AIDS, diabetes, and respiratory diseases. But the transdisciplinarity

also operates within and between specific sub-fields like molecular biology,

combinatorial chemistry, high throughput screening, genomics and bioinformatics. In

conducting knowledge exploration multidisciplinary teams of researchers are more

prominent than before. In conducting knowledge exploitation DBFs in the distinctive

sub-fields form project-based networks interacting also with ‘star’ scientists and their

teams.

As Zucker et al. (1999) see it, such projects involve no or few ‘untraded

interdependencies’, they are strictly business transactions, with contracts,

confidentiality agreements, time-limits and agreed actions (writing a patent or a paper,

for example) and outcomes. Other actors will also enter this mise-en-scène at various

points, as ‘knowledgeable attorneys’, consultants or venture capitalists (Suchman,

2000). So we may conclude that bioscientific megacentres are realised in the presence

of a nurturing ‘economic business environment’ consisting of: (1) the quality of the

inputs available to firms (e.g. human resources, physical infrastructure, availability of

information); (2) the availability and sophistication of local suppliers of components,

machinery and services, and the presence of clusters of related firms; (3) the

sophistication of local demand for advanced products and processes, including the

10

stringency of regulatory environments; and (4) the rules governing the vitality of

competition and the incentives for productive modes of rivalry (Porter, Sachs &

Warner, 2000). The key point to derive from this analysis is that while such

‘economic business environments’ are not unique to Northern California, they are far

from ubiquitous. Because they rely on massive sums of public funding, as will be

shown, a moral dilemma arises for policy makers alert to regional disparities. Should

they encourage spatial concentration to achieve global excellence or should they

encourage the development of such facilities in less favoured regions too? Keep in

mind that the latter are where negative health imbalance is often as pronounced as

economic weakness in the official statistics. Finally, as health economists and others

are coming to realise, health services and their supporting supply firms in

pharmaceuticals, biotechnology and research laboratories contribute as much as one-

sixth of GDP in some advanced economies (Cassidy, 2002).

In Table 1 data are presented on public and private health expenditure in a number of

leading OECD countries for 1997. These statistics probably underestimate the whole

Table1: Health Expenditure in Leading OECD Countries, 1997

health economy as described above by a few percentage points. Nevertheless they

demonstrate even core health expenditure running at just under 10% for the G7

countries, and much higher, at nearly 14% for the USA.

11

Statistical data for the USA at regional (State) level reveal just how skewed is the

distribution of that element of public expenditure covered by State and Federal

allocations. It should be noted that about half of US public expenditure on health is

explained as follows:

‘The system of employer-sponsored coverage emerged to restrain wage inflation during World War 2 and afterward continued when the federal courts ruled that unions could collectively bargain with employers for benefits, including health care coverage. These benefits are considered public sector ‘tax expenditures’ because they are excluded from workers’ wages for purposes of taxation and defined as an untaxed cost of business for employers’ (Milbank Memorial Fund, 2000)

Private spending occurs through private insurance taken out by individuals on top of

whatever workplace-related benefits they receive. Since the 1980s there has been a

growth in share of the former and a shift from ‘fee-for-service’ indemnity towards

‘managed care’ in insurance health plans. Health insurance has thus become more

commoditised, suppliers are keen to raise efficiencies in treatment times, use of new

technologies, and to reduce casual conduct by physicians of clinical research on

patients. Companies and consumers are equally keen to achieve value for money

under circumstances where information to enable assessment is at a premium. There

has thus been a tendency for geographical concentration of exploration research and,

due to insurance industry pressure, clinical research in specific General Clinical

Research Centres. For comparable reasons under a differently funded health insurance

system in the UK the same kind of concentration into Clinical Research Centres is

occurring (Cooke, 2002b).

In Table 2 some illustrative data are presented regarding the regional vis à vis Federal

public expenditure profile for direct health costs, excluding the employers’ ‘public

tax’ expenditures. Of particular note are the data pointing to California and New York

States accessing 11% and 12% respectively of the US total for this item. These are, of

course, large population centres, California being the larger, though it is noticeable

also that while New York spends 3.9% of its GDP in this way, California only spends

2.1%. Thus demographics, politics and per capita income also play their part in

explaining differences in level of expenditure. Among other large health expenditure

States, but not in the same rank as California and New York, are Texas and

12

Pennsylvania, followed by a group spending around $8-9 billion in 1999, namely

Florida, Illinois, Massachusetts, Michigan and Ohio. Of interest are the ways in

State State Funds ($ million) Federal Funds S/F Ratio GDP Share

New York $15,009 $14,860 99% 3.9% California $14,412 $11,981 86% 2.1% Texas $ 6,993 $ 8,100 115% 2.2% Pennsylvania $ 6,723 $ 6,110 90% 3.3% Florida $ 5,080 $ 4,469 80% 2.2% Illinois $ 5,536 $ 3,860 70% 2.1% Massachusetts $ 5,125 $ 3,279 63% 3.4% Michigan $ 4,739 $ 4,231 89% 2.9% Ohio $ 6,739 $ 1,620 24% 2.3% New Jersey $ 4,522 $ 2,971 66% 3.9% Table 2: State & Federal Health Expenditure, USA Top Ten, 1999 Source: Calculated from Milbank Memorial Fund & US Dept. of Commerce, Bureau of Economic Analysis

which such budgets are composed such that variance from the US mean GDP share

per State (2.6%) is relatively small, including the rest, where West Virginia at 4.8%

on the one hand, and Alaska at 0.6% are among the most significant outliers. With a

few exceptions, such as Ohio in Table 2, Federal funding makes a significant

contribution. Texas, like other southern Appalachian and south-western States receive

more Federal than State disbursements.

However, calculations of such disbursements in relation to State population size show

New York, at $1,572 per capita, and Massachusetts, at $1,482 to be the most generous

spenders, Pennsylvania comes next, at $1,070, then New Jersey at $949, Michigan at

$897 ahead of California at $800, followed by Illinois ($783), Texas ($765), Ohio

($763) and Florida ($615). Linking back to points made earlier about the increasing

‘commoditisation’ of health care in the US (and to a growing extent in the UK and

perhaps elsewhere) States such as these are becoming active purchasers of higher

quality, more technology- intensive, but also value for money health services. It should

further be recalled that the statistics under discussion constitute only some 20% of

total State health care budgets, but represent baseline funding. Clearly, California is

13

less dependent upon public funding than the East Coast States, that category

accounting for only just over half the GDP share it does in New York and New

Jersey. As Table 3 shows, California’s ‘mega-budget’ for private health care

expenditure easily outstrips all others among the entrants in Table 2. Hence the

State Personal Health Care Expenditures, 1998 ($mn.) California $110,057 New York $85,785 Texas $67,750 Florida $59,724 Pennsylvania $51,322 Illinois $44,305 Ohio $42,581 Michigan $35,647 New Jersey $32,695 Massachusetts $30,039 Table 3: Personal Health Care Expenditures, Top Ten US States, 1998 Source: Health Care Financing Administration (2001), Trends in State Health Care Expenditure and Funding, 1980-1998, Washington DC.

relatively low share of GDP allocated to direct public expenditure on health is

compensated for by a massive figure of some 10% of GDP being expended on

personal health care.

Keeping in mind that Table 2 represents one-fifth of each State’s health expenditure,

requiring an approximate doubling to include the employer’s contribution, and that

Table 3 accounts for some three-fifths we see the scale of total investment available

annually in the leading States’ economies. Thus California spends at least $160

billion, New York $145 billion and even modestly sized Massachusetts some $46

billion. Three key points flow from this accounting exercise. First, a few key regions,

and in fact cities in those regions have the demographic, financial and scientific scale

to afford the whole of the bioscientific and medical knowledge value chain. This

moves from the most exploratory, fundamental research into genomics and post-

genomics fields like proteomics and molecunomics. This is likely to be conducted at

specialist research institutes such as the Whitehead in Cambridge, Massachusetts

(partnered with the Sanger Institute, in Cambridge, UK and Washington University at

St. Louis for the Human Genome project). Knowledge of this kind will likely be

14

applied in specalist medical research institutes at universities, like the Dana-Farber

Cancer Institute at Harvard, or the New England Enzyme Centre at Tufts University,

Boston. Research in other key fields noted earlier (cancer, cardiovascular, AIDS,

diabetes, and respiratory diseases) will be conducted in other independent research

institutes and university research centres like Harvard Medical School’s pre-clinical

research in Biochemical & Molecular Pharmacology, Cell Biology, Genetics,

Microbial & Molecular Genetics, and Neurobiology, or affiliates like the Joslin

Diabetes Centre. Second, such research institutes and centres both attract and train

Life Sciences talent, giving critical mass to interactive research activity. This, in turn,

strongly influences growth in funding through competitive bidding to National

Institutes of Health and National Science Foundation programmes. Third, such

‘megacentres’ interact with the large hospitals, in which clinical research as well as

patient treatment occurs along with training of physicians. Massachusetts General

Hospital and the Brigham & Women’s Hospital in Boston are thus important large-

scale patient-bases for clinical trialling. There is, accordingly a suitable milieu also for

academic entrepreneurship, which, combined with Boston’s status as a top-three

location for venture capital and ‘knowledgeable attorneys’ (Suchman, 2000), makes it

a highly nurturing ‘economic business environment’ for exploitation as well as

exploration knowledge management in the form of leading biotechnology firms like

Immunex (recently acquired by Amgen), Biogen, Genzyme, Millennium,

TransKaryotic Therapies, and others, recently also joined through new openings or

acquisitions by the likes of Abbott Laboratories, AstraZeneca, Aventis, Pfizer and

Wyeth (recently merged with American Home Products).

Thus, while scale of expenditure in general health systems clearly matters as an entry

ticket to the megacentre ‘tournament’, it is by no means a sufficient condition. There

has also to be world-class science and world-class commercialisation capability.

There has to be localised ‘social capital’ among the actors present, which can link

appropriate partners across epistemic community boundaries. Firms help themselves

when they speak with a single voice on matters of common concern, something

portrayed in the Boston cluster by the activities of the 280-member Massachusetts

Biotechnology Council. Much of this system is portrayed in Fig. 1 below.

Fig. 1: The Cambridge, Massachusetts and Greater Boston Biosciences Megacentre

4. Rational Drug Design & DBF Networks

What are the difficulties faced by DBFs in taking advantage of such nurturing

economic business environments, and in what ways are those in the USA better

placed to benefit than those in Europe? One key feature that differentiates them is the

estimated $20 billion per year that has been available for US biotechnology research

from Federal investment in, for example, the 1994-98 period studied by Senker &

Van Zwanenberg (2001). This compares with the approximately €10 billion spent by

European governments over the same period. Further, it is argued, US DBFs can

exploit the research findings of National Institutes of Health-funded research more

swiftly and efficiently due to the existence of National Science Foundation sponsored

Small Business Innovation Research grants enabling DBFs to develop ideas more

quickly thus potentially influencing venture capitalists and ‘big pharma’ to invest in

what elsewhere would appear to be more high-risk ventures. These SBIR grants arise

from a requirement that R&D spending Federal government departments must spend

up to 2.5% of their extra-mural research budgets on commissions from SMEs.

Moreover, research-minded entrepreneurs requiring continuing interaction with other

discovery firms or research institutes have more of these to choose among and thus

exploit better networking opportunities.

Jaffe et al.’s (1993) finding that knowledge spillovers from universities to firms were

relatively regional or even local was found to be true in the Senker & Van

Zwanenberg research on European biopharmaceutical firms. Exacerbating this, Owen-

Smith et al. (2001) found that public research organisations (PROs), with which such

DBFs are likely to seek to interact, have far more intensive and extensive inter-

institutional research networks than European firms. The latter had much smaller,

sometimes dyadic, networks and these were often nationally constricted, except for

occasional transatlantic contact with a US institute, and consequent potential linkage

into wider US knowledge networks. However, so few and restricted were the active

international linkages that less knowledge exploitation was feasible, and more slowly,

than in the case of the better-networked US institutes. Hence, it can be argued from a

17

‘capabilities’ perspective as outlined in Section 1 of this paper, that the better

networking for purposes of knowledge exploration and support for knowledge

exploitation displayed by US research institutes is a sign of superior institutional

capability. Given their role as key intermediaries in the knowledge management

process, the institutional capability in the US Biosciences Innovation System, adds

significant value to an already doubly advantaged initial research resource base.

It can be further shown, as Owen-Smith et al. (2001) go on to do, that US networks

among PROs are hierarchically structured. Over the period from 1988-1998, the

Boston cluster has remained at the peak of US interorganisational research linkages,

and its connections to other clusters has doubled to over 50% of its total contacts over

the period. San Diego has moved to second place in the hierarchy over San Francisco,

while Seattle and New York have exchanged positions, Seattle rising above New

York in terms of the number and strength of its inter-regional PRO linkages. Thus

relationships between PROs and firms in their clusters are augmented by the

institutional capabilities of both to benefit from spanning therapeutic areas, engaging

in multiple stages of the knowledge exploitation value chain, and involving diverse

collaboration. The institutional system takes on certain characteristics that have

resulted in the term ‘collaboratory’ being used to describe such interorganisational

networking. As has been suggested, European DBFs and PROs have tended to engage

in more attenuated innovation networks, with more specialised than diverse

interactions, and a more limited external value chain involvement that is also mostly

national in character.

The weaknesses of European international networks of PROs and DBFs can be

understood in terms of the relative immaturity of regional science infrastructures. This

in turn, echoes the relatively small budgets that have traditionally been available for

systemic medical, health and life sciences research integration. This is changing

rapidly in some European countries, and has been rather better developed in some

smaller EU economies for some time. Thus Germany and France have changed

business regulations to encourage innovation and academic entrepreneurship, while a

long period of under funding in the UK health service has been reversed both in

service delivery budgets and scientific research budgets. In the UK, the world’s

18

largest scientific research charity, the Wellcome Trust, has become highly active in

co-funding with the UK government investments in research infrastructure and basic

exploration research such as the Human Genome and new post-genomic research at

such centres of excellence as the Sanger Research Institute at Cambridge (UK).

Nevertheless, some important parts of the exploration and exp loitation infrastructure

remain underdeveloped to varying degrees. In the UK, much emphasis has been

placed upon funding Centres of Excellence. These are relatively few in number and

highly imbalanced between regions and within them. In an earlier paper (Cooke,

2002b) it was shown that implicitly or explicitly policies are set to strengthen specific,

often potentially or actually multi-purpose, sites with the strongest possibility for

elaborating a medical and life sciences knowledge value chain to the fullest extent.

Modelled to some degree on Boston, Cambridge (UK) has been a major recipient of

Wellcome Trust and UK research council funding, complemented by special Treasury

funding to enhance academic entrepreneurship through linkage with the

Massachusetts Institute of Technology, and new investments in hospital

infrastructure, including a Clinical Research Centre, to fill out its emergent

megacentre character. It is well known that Cambridge was already the UK’s leading

biotechnology commercialisation location, consequent on long term cultural change

efforts to stimulate academic exploitation of world class exploration knowledge. A

case in point of the latter is the shift in the Medical Research Council’s Molecular

Biology Institute from essentially giving away Intellectual Property Rights to

discoveries like Monoclonal Antibodies as occurred with Milstein and Kohler’s

discovery, to stimulus for contemporary scientists to become academic entrepreneurs.

Now the MRC lab has some 30 spinout firms, some like Cambridge Antibody

Technologies valued in many millions of pounds on the UK’s technology stock

market.

Yet this process can often be rather unsystematic, as the experience in neighbouring

Oxford, also home to world- leading medical and life sciences research, and to

numerous start-up and more mature biotechnology firms like Oxford Glycosciences,

Oxford Asymmetry and Xenova. A key part of the maturation of exploration into

exploitation knowledge occurs in incubators. Oxford Innovation, under the wing of

the Oxford Trust, a non-profit commercialisation institution, runs Oxford

19

BioTechNet, responsible for a number of early and mid-stage start-up biotechnology

firms. However, such is the relatively precarious nature of the funding needed for

such businesses that, in reality, it is impossible to posses all the skills required of

intermediaries such as those found in US megacentres like Boston in one incubator or

even one small university city. Thus issues of funding, from seedcorn to business

angel to full-blown venture capital, legal questions, ranging from incorporation to

patent defence, management issues like in- licensing and out- licensing, development

of supply chain linkages or milestone payment agreements with ‘big pharma’, and

more mundane questions of property acquisition as companies grow and leave the

incubator, are all dealt with by a single incubator manager.

Interestingly, through a research partnership with such well- found innovation

intermediaries as BioM in Munich and GenoPole in Paris, it is clear that comparable

difficulties apply in their cases too. That is, finance may not be such a problem as

finding appropriate expertise. For even though it is said that Munich had some thirty

venture capitalists in the seven years after BioRegio, the German Federal

government’s regional biotechnology commercialisation initiative, such is the

newness of most of the twenty or so new biotechnology firms that special early-stage

management expertise and support is a greater weakness than investment finance

(Kaiser, forthcoming). As will be seen below, Munich is now assessed as being

Europe’s leading high technology cluster, at least in ITC if not biotechnology. But

while it has a few mature biotechnology firms kike MediGene and MorphoSys, most

of its start-ups are in their infancy, heavily dependent on public venture finance, and it

is unclear how viable many are were they to be wholly reliant on even the protected

regime by which most biotechnology firms survive without showing profitability.

This compares rather unfavourably with the picture of relatively healthy research- led

networking painted by Orsenigo et al (2001) where collaborative relationships among

firms and research institutions initiated from the US are shown to be diverse, capable

of including linkage with some of the above-mentioned European firms, and

sophisticated enough in knowledge management of the revolutionised ‘rational drug

design’ methodologies consequent upon the revolution in molecular biology

(Henderson et al., 1999) for ‘big pharma’ to seek entry to the networks rather than

20

vice-versa. The change from a ‘chance discovery’ model of scientific research to a

‘rational drug design’ model based on combinatorial chemistry, molecular biology,

high throughput screening, genomics and bioinformatics has meant that those regions

and localities with clusters of DBFs of various kinds, linked also to ICT firms and

knowledge management intermediaries are absolutely advantaged, even to the extent

of making ‘big pharma’ dependent on them for key knowledge of both the exploration

and exploitation kind. Increasingly such DBFs even manage the due diligence and

trial management processes, leaving ‘big pharma’ to exchange contracts with the DBF

networks for the license to market and distribute the hoped-for biopharmaceutical

drug at the end of the pipeline.

5. Which Are Leading Megacentres and What Is Regional Science Policy?

We have seen already that Boston is perhaps the leading biosciences megacentre, not

because it has the heaviest medical or even bioscientific research budgets, but because

it is presently one of, if not the leading centre for exploration research. So much so

that while the Swiss drug company Novartis announced in 2000 a path-breaking

agreement to spend $25 million on first access to the results of plant and microbial

biology research conducted at the University of California, Berkeley, in the heart of

the Northern California biotechnology cluster, in 2002 Novartis announced the

establishment of a $250 million Novartis Genomics Research Institute in Cambridge,

Massachusetts on the grounds that it was the leading exploration and exploitation

centre for genomics and post-genomics knowledge. Boston’s current primacy has not

been the product of the operations of the market mechanism alone. In 1999, $770

million of mainly public or charitable research funding was earned for medical and

bioscientific research. That figure is likely to have exceeded $1 billion shortly

afterwards. This was marginally less than the amount of National Institutes of Health

funding alone passing through the Northern California cluster in 1999, a statistic that

increased to $893 million in 2000 (CHI/PWC, 2002). Most of the exploration research

conducted in both Cambridge/Boston and Northern California is conducted in

institut ions that are dependent on public funding, though private research foundations

are also functional in both. In Boston, the Massachusetts Biotechnology Council is an

active and successful biotechnology association that lobbies industry and political

forums at State and Federal levels, pressing for an FDA presence in Boston to offset

21

the advantage enjoyed by emergent firms and research institutes located in Maryland

near the head offices of both NIH and FDA (see table 4).

Institution Rank (1994) Amount ($million) Program Resources Inc., Reston VA 1 $98.0 Westat Inc., Rockville, MD 2 $50.0 Adv. Biosc. Lab Inc, Kensington, MD 3 $30.6 U. of Alabama, Birmingham, AL 4 $16.2 Research Triangle Institute, RTC Park, NC 5 $15.1 Johns Hopkins U., Baltimore, MD 6 $14.6 ROW Sciences Inc, Rockville, MD 7 $14.5 Harvard U., Cambridge, MA 8 $13.2 Southern Research institute, Birmingham, AL 9 $12.9 U. of Texas Health Science Centre, Houston, TX 10 $11.3 Table 4: Top Ten Institutions for NIH R&D Contracts, 1994 Source NIH

The figures in Table 4 are somewhat out of date but their significance lies in the

evidence of a few years ago that geographical proximity explained the largest share in

the variance of grant allocations by NIH, something that has been changed somewhat

by the more piecemeal evidence that Johns Hopkins and Harvard Universities now vie

for top position in consequence of their recognition of various biases in the system

revealed by the statistics for 1994’s allocations.

Let us look more closely at the manner in which the assets of biosciences megacentres

like those in Northern and Southern California are now packaged in documentation

that promotes the image that is intended to appeal to investors of all kinds into the

regional innovation system. The two following examples, from Northern and

Southern California are produced by a non-profit association (The California

Healthcare Institute) and a consultancy (Michael Porter’s Monitor) respectively. The

California Healthcare Institute is a public policy institute for California’s 200 leading

biotechnology firms and research institutes. It is thus comparable to the Massachusetts

Biotechnology Council. Its political brief is expressed clearly by CEO Gollaher who

despite noting ‘…funding for basic science is strong..’ bemoans the fact that ‘…many

federal and state lawmakers advocate policies that would impede medical innovation.

Our greatest threats include a total ban on human cloning ad severe restrictions on

stem cell research; a Medicare administration that…. effectively excludes new

22

products…. and a leaderless FDA facing the greatest wave of new inventions in

history’ (Gollaher in CHI/PWC, 2002). The demand is for collaboration among

members of the biosciences innovation system to change laws that are perceived as

threatening the evolution of the industry in the post-genomic era.

Northern California is presented as the birthplace of biotechnology, which, along with

biomedical innovatons like cardiac stents, has a strong base of some 819

biomedical/biotechnological firms, employing 86,000 people (28,000 in

biotechnology), total R&D of $1.1 billion, NIH grants of $893 million and $4.1

billion in worldwide revenues, including $2.7 billion exports.. New infrastructure

projects inc lude University of California San Francisco Medical School’s new $1.4

billion Mission Bay bioscience research campus, the new California Institutes for

Science and Innovation, and the California State University CSUPERB joint ventures

programme where universities and the private sector collaborate in bioscience

research, technology transfer, business and even residential development. Much

emphasis is placed on survey results showing that significant interaction occurs

among firms and the institutional research base in Northern California.

Thus, California’s academic research institutions are credited with playing a central

role in the growth of nearly one-third of biomedical/biotechnological firms, 42% of

firms had at least one research contract with a California research institution, 56% of

firms planned to broaden or maintain such agreements and up to 70% of firms having

patent license agreements planned to maintain or broaden them in future. The key

Northern California life sciences and clinical research institutes cited include Stanford

University (Biomedical Technology Information Programme), Lawrence Berkeley

and Lawrence Livermore National Laboratories, and the University of California, San

Francisco Medical School, Berkeley (BioSTAR industry-academic collaboration),

Santa Cruz (with Berkeley, the California Institute for Bioengineering, Biotechnology

& Quantitative Biomedical Research, QB3) and Davis (Life Sciences Information

programme). More than 19,000 are employed in research in the region, and nowadays

two of the top ten NIH R&D grant recipients in the US are UCSF and Stanford. A

picture of this cluster that is developing the characteristics of a biosciences

megacentre is presented in www.biospace.com. But the judgement as to whether it yet

is, is occluded by the lobbying points that although some larger firms such as Abbot

23

Laboratories and Genentech are present, most are SMEs, 49% without products on the

market, 45% with no revenue in 2000. Finally, of the pharmaceutical pipeline

products reported, 53% are in preclinical trials. This is by no means unusual, but

nevertheless testifies to the apparent fragility of the exploitation aspect of the

Northern California cluster, once its strength, but never adequately backed up with

strong bioscience exploration capabilities and now, belatedly perhaps, seeking to

embed them.

In Southern California, the San Diego biotechnology cluster has larger claims to be

considered a biosciences megacentre than even that in the North. In Porter’s (2002)

competitiveness study San Diego’s biopharmaceuticals cluster is presented as long

established and among the most significant outside Boston, especially for R&D.

Cluster employment growth was more than 8,000 from 1988-1997 and San Diego had

the most rapid growth in patent output compared to the twenty largest US

biotechnology clusters. There are some 400 SMEs, focusing mainly on one or two

preferred drug targets, the University of California, San Diego, with numerous

specialist research centres, and finally, some globally known research institutes, the

Salk Institute, the Scripps Research Institute, the Burnham Institute, and the La Jolla

Institute for Allergies and Immunology, each focusing upon aspects of life science,

medical or clinical research. The Scripps Institute, since establishment in the 1950s

required its researchers to raise their own funds encouraging collaborative innovation

with larger firms (like Dow). By contrast, the Salk Institute does not conduct

corporate research but licenses its discoveries and takes equity stakes in companies.

UCSD emphasised medical research and academic entrepreneurship. One early fruit

of that approach was Hybritech, a 1978 biotechnology start-up, from which more than

fifty other local biotechnology DBFs were spun out. In 1986 it was sold to Eli Lilly

for $400 million. A further feature of this cluster is its strong and long-established

networking propensity, signified by the establishment since 1985 of the UCSD

CONNECT network, a model for cluster integration in many other new economy

clusters such as Scotland’s and Cambridge’s (UK) (ICT) networking associations.

In Fig. 2 an analysis is provided of the origins of the San Diego biotechnology cluster

firms in relation to the San Diego CONNECT network. It is notable that

24

‘entrepreneurs’ (like the first owners of Hybritech) are far more important sources of

direct spin-out firms than either the university or research institutes.

`

CONNECT (San Diego) Biotechnology Cluster Linkages (Adapted from Lee, C. & Walshok, M., 2001, Making Connections, Report to UC Office of the President)

25

Overall, the Lee and Walshok (2001) report concludes that the San Diego

biotechnology cluster is characterised by the following features:

• 400+ Companies

• 248 emergent firms

• 28,000 employees

• UCSD CONNECT – ‘ a network of professional competencies focused on building shared knowledge…. for technological companies’

• UCSD, Salk Institute, Scripps Research Institute, Burnham & La Jolla Institutes

• an innovation support infrastructure of investors, consultants and technology intermediaries

The judgement of Porter’s (2002) team is that the close proximity of research centres

and firms on the Torrey Pines Mesa was a key advantage in encouraging collaboration

and growth. Regarding patenting San Diego registered 360 patents in

biopharmaceuticals in 1997, a rate of 13.17 per 1,000 workers and the growth rate

was the US’s fastest, though the intensity was less than nine other bioclusters.

Venture capital was invested at a much higher rate than nationally with $421 million

having been placed 1995-99, nearly 10% of the national total. But research

organisations are the greatest strength, with Novartis and Dow having joined the

public institutes, making a total of some 16,000 employees in biopharmaceuticals

research alone, larger than that specific category in Northern California. However,

both Californian clusters have strongly emergent megacentre properties, based

especially on their strength in exploration knowledge and an abundance of SME

DBFs that are capable of rapidly forming molecular discovery networks due to

geographical proximity and critical mass. In the Orsenigo et al. (2001) study of

rational drug design research networks18% of interacting firms and institutes were in

Boston, 16% in Northern California and 12% in San Diego. The few partners outside

the USA were located in Munich (2), Cambridge (2) and Oxford (2).

This brings us neatly to brief consideration of the status of bioscience megacentres in

Europe. In terms of possession of the key exploration and exploitation institutes and

firms, it is clear from maps produced regularly by Ernst & Young, e.g. (1999) that the

greater London area has the greatest number of DBFs concentrated in an

agglomeration with few exploration institutions south and west of London and two

clusters at Cambridge and, slightly smaller, Oxford whe re exploration and exploitation

26

go hand in hand in geographical proximity. Cambridge has the strongest case for being

considered a potential megacentre at present. Cambridge has a rather diverse

biotechnology processing and development as well as services support structure, even

though the industry is relatively young and small. Some of the service infrastructure

and perhaps the equipment sector benefits from the earlier development of Information

Technology businesses, many also spinning out from university research in

Cambridge. The infrastructure support for biotechnology in and around Cambridge is

impressive, much of it deriving from the university and hospital research facilities.

The Laboratory of Molecular Biology at Addenbrookes Hospital, funded by the

Medical Research Council; Cambridge University’s Institute of Biotechnology,

Department of Genetics and Centre for Protein Engineering; the Babraham Institute

and Sanger Institute with their emphasis on functional genomics research and the

Babraham and St. John’s incubators for biotechnology start-ups and

commercialisation, are all globally-recognised facilities, particularly in

biopharmaceuticals. However, in the region are also located important research

institutes in the field of agricultural and food biotechnology, such as the Institute for

Food Research, John Innes Centre, Institute of Arable Crop Research and National

Institute of Arable Botany. Its core biotechnology industry consists of approximately

90 firms and the broader cluster (venture capitalists, patent lawyers etc.) consists of

approximately 200 firms, with the core biotechnology firms employing 2,500-3,000

people. Of relevance to the biotechnology community are the activities of the Eastern

Region Biotechnology Initiative. This biotechnology association is the main regional

network with formal responsibilities for: newsletters, organising network meetings;

running an international conference; web-site; sourcebook and database on the

bioscience industry; providing aftercare services for bio-businesses; making intra- and

inter-national links (e.g. Oxford, Boston, San Diego); organising common purchasing;

business planning seminars; and government and grant-related interactions for firms.

Munich seems to have become a more significant biotechnology player in recent

times, particularly since the onset of BioRegio in 1995. It was well established in

exploration knowledge institutions but like the rest of Germany, weak in exploitation

mechanisms. This seems to have been re-balanced but it is too soon to say how

significantly as key second-round funding demands are only in 2002 coming on-

stream. The science base in Munich is broad, but with special expertise in health-

27

related and agricultural and food biotechnology. There are three Max Planck Institutes

of relevance, in Biochemistry, Psychiatry and the MPI Patent Agency. GSF is the

Helmholtz Research Centre for Environment and Health, and the German Research

Institute for Food Chemistry is a Leibniz Institute. There are three Fraunhofer

Institutes, one of Germany’s four Gene Centres, two universities and two

polytechnics. The main research-oriented ‘big pharma’ companies are Roche

Diagnostics (formerly Boerhinger Mannheim) and Hoechst Marion Roussel (since

2000, Aventis). The work areas of this science community include; 3D structural

analysis, biosensors, genomics, proteomics, combinatorial chemistry, gene transfer

technologies, vaccines, bioinformatics, genetic engineering, DNA methods, primary

and cell cultures, microrganisms, proteins, enzymes and gene mapping. The Bavarian

commitment to biotechnology (and other new technologies) was realised through its

state government decision to privatise its share in power-generation and distribution

companies in the 1990s, thereby creating a funding pool to subsidise applied

technology developments. Nevertheless, the numbers of start-ups are not

overwhelming, perhaps because of the quest for ‘quality’ start-ups in whom

substantial sums may be individually invested.

A further explanation for conservatism is that BioM AG, the ‘midwife’ agency to the

cluster, funded partly by BioRegio and set up as a corporation, makes investments

with its shareholders’ (state, industry and banks) money. Banks seeking to earn high

returns hold most of the shares. They also use this method to learn about

biotechnology, its risks and prospects. Thus, an already well-subsidised system is

further protected from risk by the influence of banking culture, itself highly

conservative in Germany, to ensure, as far as possible, risk avoidance. Hence, while

BioM is the network face of the biotechnology cluster in Munich, its activities are

ultimately orchestrated indirectly and directly by the banks, abetted by a fairly risk-

averse, mostly publicly-funded, venture capital industry and the local pharmaceutical

and chemical companies (Giesecke, 2000).

So, for the moment Europe’s best candidates (to which may possibly be added

Stockholm-Uppsala in Sweden, VINNOVA, 2001) are either somewhat lacking in the

maturity or critical mass of their DBFs, as is the case to varying degrees of both

Cambridge and Munich. This is an exploitation rather than exploration knowledge

28

deficiency. It is arguable, judged by Nobel Prizes awarded that Cambridge with

eleven is superior in biosciences to any of the stronger candidates in the USA. But it

is unquestionable that the latter have been more entrepreneurial, even though large

numbers of DBFs everywhere have neither products nor profits.

What does this signify? In an earlier paper, an extensive analysis of the response by

the large number of non-megacentre, even non-bioscience regions in the USA,

Canada and the UK was undertaken (Cooke, 2002b). This showed that fifteen US

States had undertaken science base analyses and developed science strategies with

targets and mechanisms for augmenting their capture of basic science funding in

biosciences. Some piecemeal evidence of the outcomes of these are the statistics that

show universities like Harvard, Johns Hopkins, UCSF, Stanford and Duke as

occupying highest places in the NIH R&D allocations compared to hitherto. But the

US has also an interesting mechanism for supporting ambitious if underdeveloped

universities in lagging States to access national competitive funding. The EPSCOR

scheme under the NSF competitive bidding procedure for accessing, first programme,

now project funding allows designated States, e.g. Oklahoma and Mississippi to bid

competitively for such bioscience research funding by lowering the grant aid

qualifying bar below that expected of the rest of the USA. This, in EU terms, is a

‘structural funds’ mechanism within the Framework Funds and it has produced

generally positive rather than futile results in the USA.

In Europe, some regional bodies have begun to develop regional science strategies,

most notably Scotland that identifies its £800 million R&D spend annually and seeks

to augment it through intra-regional collaboration and the furtherance of existing

science-funding mechanisms. Medical and Biosciences are two if the three funding

areas to be targeted. In Finland a central government policy of seeding regional

Centres of Expertise, and later, Centres of Excellence has produced up to six regional

bioscientific Centres of Excellence that follow the US system of linking exploration

to exploitation but through largely public rather than private knowledge exploitation

mechanisms. Regional governance institutions are now somewhat more aware now

than they were until only very recently, that health is a large part of their economy,

that it links directly to some of the most exciting science being done in the world

today, and that as well as making a direct economic contribution through purchasing

29

and employment, it can make an important indirect contribution to economic welfare

through possible academic and corporate spin-out activities. It is likely that regional

analysis and policy will be more rather than less influenced by issues such as those

discussed in this paper in future.

6. Conclusions

These are brief and orientated towards future implications for regional science and

policy rather than being a simple reprise of what has been said. First, of significance

to the never-ending debate about the viability of SMEs in a globalising world

dominated by multinational firms, events during the past decade or so show just how

misplaced arguments prophesying the demise of SME significance can be. It is ‘big

pharma’ that is in crisis as its traditional expertise in fine chemistry is subverted by

the molecular biology revolution and the demand for transdisciplinary teams of DBFs

to form project-based networks to seek ‘rational drug design’ solutions based on

reagents and their inhibitor compounds at the molecular and even sub-molecular

levels. Meanwhile ‘big pharma’s’ drug innovation pipeline dries up as R&D costs

escalate, giving a further imperative to externalise projects to the ‘knowledge value

chain’. DBFs in turn rely effectively upon the deep pockets of ‘big pharma’ for

licensing cash, marketing and distribution. Some DBFs are now quite large and one

(Celltech) recently bought a mid-size ‘pharma’ (Medeva) in the UK. Other global

‘pharma’ has been merging at pace recently as they seek to reap one-off shareholder

value from reducing the competition as Glaxo, Pfizer, Wyeth, Aventis and Novartis to

name a few, testify. The regional science implications of these shifts towards public

R&D exploration and exploitation and away from ‘big pharma’s’ traditional

metropolitan redoubts require swift and serious analysis.

In regional policy terms, the new model is more foresight-driven, more collaborative,

based on shared vision and leadership than old redistributive policy used to be. This is

necessarily policy for securing ‘generative growth’ (Cooke, 2002c). It must take

seriously the long-established presence of hospitals and universities and seek to forge

links along the biosciences value chain. It must do this in a sometimes hostile

environment in which national or federal governments prefer to see a few Centres of

Excellence, possibly not too far away from their seats of government, rather than in

30

obscure and peripheral regions. Yet policy-makers in precisely such regions will

surely seize upon the obvious notion that for once in history public sector investments

are new sources of innovative development and generative growth, seeking ways of

mobilising enterprising coalitions to bid for infrastructure and ‘star’ scientist

recruitment from a variety of funding sources. Or, if this is insufficient, pressing their

multi- level governance structures for affirmative research allocation action such as

that already pioneered in the USA through EPSCOR.

References

Audretsch, D. & Feldman, M. (1996) Knowledge spillovers and the geography of

innovation and production, American Economic Review, 86,630-640

Becattini, G. (1989) Sectors or districts: some remarks on the conceptual foundations

of industrial economics, in E. Goodman & J. Bamford (eds.) Small Firms &

Industrial Districts in Italy, London, Routledge

Best, M. (2001) The New Competitive Advantage, Oxford UP

Brusco, S. (1989) A policy for industrial districts, in E. Goodman & J. Bamford (eds.)

Small Firms & Industrial Districts in Italy, London, Routledge

California Healthcare Institute & PriceWaterhouseCoopers (2002) Biomedicine: the

New Pillar in Northern California’s Economy, La Jolla & San José,

CHI/PWC

Chandler, A. (1990) Scale & Scope , Cambridge, Harvard/Belknap

Coase, R. (1937) The nature of the firm, Economica, 4, 386-405

Cooke, P. (2002a) Knowledge Economies, London, Routledge

Cooke, P. (2002b) Towards regional science policy? The rationale from biosciences,

paper to Conference on ‘Rethinking Science Policy: Analytical Frameworks

for Evidence-Based Policy’, Science Policy Research Unit, University of

Sussex, March 21-23

Cooke, P. (2002c) Future challenges for regional development posed by economic

globalisation, International Journal of Technology Management

(forthcoming)

Curzio, A. & Fortis, M. (eds.) (2002) Complexity & Industrial Clusters ,

Heidelberg, Springer

31

Dosi, G. (1988) Sources, procedures and microeconomic effects of innovation,

Journal of Economic Literature , 26, 1120-1171

Ernst & Young plc (1999) European Life Sciences: Sixth Annual Report, Reading,

Ernst & Young

Feldman, M. & Audretsch, D. (1999) Innovation in cities: science-based diversity,

specialisation and localised competition, European Economic Review, 43,

409-429

Galison, P. (1997) Image and Logic: A Material Culture of Microphysics, Chicago

UP

Gibbons, M. et al. (1994) The New Production of Knowledge, London, Sage

Giesecke, S. (2000) The contrasting roles of government in the development of

biotechnology industry in the US and Germany, Research Policy, 29, 205-223

Glaeser, E., Kallall, H., Scheinkman, J, & Shleifer, A. (1992) Growth in cities,

Journal of Political Economy, 100, 1126-1152

Griliches, Z. (1992) The search for R&D spillovers, Scandinavian Journal of

Economics, 94, 29-47

Henderson, R, Orsenigo, L. & Pisano, G. (1999) The pharmaceutical industry and the

revolution in molecular biology: interactions among scientific, institutional

and organisational change, in D. Mowery & R. Nelson (eds.) Sources of

Industrial Leadership, Cambridge UP

Jacobs, J. (1969) The Economy of Cities, New York, Random House

Jaffe, A, Trajtenberg, M. & Henderson, R. (1993) Geographic localization of

knowledge spillovers as evidenced by patent citations, Quarterly Journal of

Economics, 108, 577-598

Kaiser, R. (forthcoming) Innovation policy in a multi- level governance system: the

changing institutional environment for the establishment of science-based

industries, European Planning Studies, 11

Lee, C. & Walshok, M., (2001) Making Connections , Report to University of

California Office of the President

Marshall, A. (1918) Industry & Trade , London, Macmillan

Milbank Memorial Fund (2000) State Healthcare Expenditures, New York,

Milbank Memorial Fund

32

Niosi, J. (2002) Canadian industry clusters: aerospace, biotechnology and software,

Paper to Innovation Systems Research Network Conference, Quebec City,

May 7-9

Nonaka, I. & Takeuchi, H. (1995) The Knowledge-Creating Company, Oxford UP

Orsenigo, L, Pammolli, F. & Riccaboni, M. (2001) Technological change and

network dynamics: lessons from the pharmaceutical industry, Research

Policy, 30, 485-508

Owen-Smith, J, Riccaboni, M, Pammolli, F. & Powell, W. (2001) A comparison of

US and European university- industry relations in life sciences, (mimeo)

Penrose, E. (1959) The Theory of the Growth of the Firm, Oxford UP

Polanyi, M. (1966) The Tacit Dimension, London, Routledge

Porter, M. (1990) The Competitive Advantage of Nations , New York, The Free

Press

Porter, M. (1998) On Competition, Boston, Harvard Business School Press

Porter, M. Sachs, J. & Warner, J. (2000)

Porter, M. (2002) Clusters of Innovation: regional Foundations of US

Competitiveness, Washington, Council on Competitiveness

Piore, M. & Sabel, C. (1984) The Second Industrial Divide , New York, Basic Books

Richardson, G. (1972) The organisation of industry, Economic Journal, 82, 883-896

Sabel, C. (2002) Diversity, not specialisation: the ties that bind the (new) industrial

districts, in Curzio, A. & Fortis, M. (eds.) Complexity & Industrial Clusters ,

Heidelberg, Springer

Seely Brown, J. & Duguid, P. (2000) The Social Life of Information, Boston,

Harvard Business School Books

Senker, J. & Van Zwanenberg, P. (2001) European Biotechnology Innovation

Systems , Final Report to EU-TSER Programme, University of Sussex, SPRU

Stankiewicz, R. (2002) The evolving design space of technology and the public R&D

system, paper to Conference on ‘Rethinking Science Policy: Analytical

Frameworks for Evidence-Based Policy’, Science Policy Research Unit,

University of Sussex, March 21-23

Stewart, T. (2001) The Wealth of Knowledge, London, Nicholas Brealey

Suchman, M. (2000) Dealmakers and counsellors: law firms as intermediaries in the

development of Silicon Valley, in M. Kenney (ed.) Understanding Silicon

Valley, Stanford UP

33

Teece, D. & Pisano, G. (1998) The dynamic capabilities of firms: an introduction, in

G. Dosi, D. Teece & J. Chytry (eds.) Technology, Organisation and

Competitiveness, Oxford UP

VINNOVA (2001) The Swedish Biotechnology Innovation System, Stockholm,

VINNOVA

Williamson, O. (1985) The Economic Institutions of Capitalism, New York, The

Free Press.

Zucker, L, Darby, M. & Armstrong, J. (19980 Geographically localised knowledge:

spillovers or markets? Economic Inquiry, 36, 65-86